A system was developed for exposure of unanesthetized mice to airborne chemicals and for continuous measurement of their breathing pattern prior to, during and following exposure. By measuring inspiratory and expiratory airflows (VI and VE), and integration with time to yield tidal volume (VT), we obtained characteristic modifications to the normal breathing pattern. These permitted recognition that a specific portion of the respiratory tract was affected by the selected airborne chemicals. Following recognition, we also quantitated the degree of effect using one specific measurement in each case. An effect on the upper respiratory tract, induced by the sensory irritant, 2-chlorobenzylchloride, was quantitated by measuring a decrease in respiratory frequency. An effect on the conducting airways, induced by the airway constrictor, carbamylcholine, was quantitated by a decrease in VE at the mid-point of VT. An effect at the alveolar level, induced either by the vagal nerve ending stimulant, propranolol, or by the pulmonary irritant, machining fluid G, was quantitated by an increase in the length of a pause induced at the end of expiration. The system is easy to construct and operate and can be used to rapidly evaluate the effects of airborne chemicals on the respiratory tract.

The pattern and timing of a normal breath in unanesthetized mice was analyzed from measurement of inspiratory and expiratory airflows (VI and VE). Airflow was measured via a differential pressure transducer, attached to a pneumotachograph, which itself was attached to a body plethysmograph into which a mouse was placed. The analog voltage from the differential pressure transducer was digitized and stored for analysis on a microcomputer. Criteria were developed to classify each breath as normal (N) or belonging into one of seven abnormal categories. The abnormal categories were arrived at by computer analysis, recognizing specific modifications of the normal pattern into patterns of: sensory irritation of the upper respiratory tract (S), airflow limitation within the conducting airways of the lungs (A) or pulmonary irritation at the alveolar level (P). Combinations of these effects, i.e., S+A, P+A, P+S and P+S+A were also recognized. Computer analysis of each breath also permitted quantitative evaluation of the degree of S, A or P abnormalities. To induce each type of effect we used inhalation exposures to 2-chlorobenzylchloride, carbamylcholine or propranolol. We propose that this approach will permit rapid evaluation of the possible effects of airborne chemicals at three levels of the respiratory tract, with the classification of the type of effect easily obtained in an objective way using well defined criteria, followed by quantitation of the degree of each effect.

A database was developed for chemicals whose sensory-irritating properties had been investigated using a previously described animal bioassay. In this bioassay, mice were exposed to an airborne chemical, and changes in their respiratory pattern were determined. For each chemical tested, the concentration capable of producing a 50% decrease in respiratory rate (RD50) was obtained and its relative potency estimated. For the current study, 295 such airborne materials, including single chemicals and mixtures, were found in the literature. A total of 154 RD50 values were obtained in male mice of various strains for the 89 chemicals in the database for which there were also TLVs. An examination of the TLVs and RD50 values demonstrated, as previously with the smaller dataset (n = 40), a high correlation (R2 = 0.78) of the TLVs with 0.03 x RD50. This supports the continued use of the animal bioassay for establishing exposure limits to prevent sensory irritation in the workplace. No other bioassay provides this type of information or has been used so extensively to suggest guidelines for occupational exposures.

We used a database of 145 volatile organic chemicals for which the sensory irritation potency (RD50) has been reported in mice. Chemicals were first separated into two groups: nonreactive and reactive, using Ferguson's rule. This rule suggests that nonreactive chemicals induce their effect via a physical (p) mechanism (i.e., weak forces or interactions between a chemical and a biological receptor). Therefore, appropriate physicochemical descriptors can be used to estimate their potency. For reactives, a chemical (c) mechanism (i.e., covalent bonding with the receptor) would explain their potency. All chemicals were also separated on the basis of functional groups and subgroups into 24 classifications. Our results indicated that the potency of nonreactive chemicals, regardless of their chemical structure, can be estimated using a variety of physicochemical descriptors. For reactive chemicals, we identified five basic reactivity mechanisms which explained why their potency was higher than that estimated from physicochemical descriptors. We concluded that Ferguson's proposed rule is adequate initially to classify two separate mechanisms of receptor interactions, p vs c. Several physicochemical descriptors can be used to estimate the potency of p chemicals, but chemical reactivity descriptors are needed to estimate the potency for c chemicals. At present, this is the largest database for nonreactive-reactive chemicals in toxicology. Because of the wide variety of c chemicals presented, a semi-quantitative estimate of the potency of new, or not previously evaluated, c chemicals can be arrived at via comparison with those presented and the basic chemical reactivity mechanisms presented.

No abstract